Study On The Stability And Quality Of Thin-walled Workpiece In High-speed Milling

Thin-walled parts are widely used in aerospace and defense industrial production.It is,likely to occur deformation and vibration during processes because of its complex structure,large machining allowance,poor overall rigidity and poor processing manufacturability.Reaserch on high speed milling of thin-walled parts has important theoretical and practical value.According to application background and main technical problems for thin-walled parts,based on the high-speed milling technology,around stability analysis,deformation prediction and error compensation,surface quality control,processing efficiency technology and other issues,systematic study is conducted in this paper.Through the analysis of the milling process,considering the influence of the regenerative chatter on the stability,one-dimensional,two-dimensional,three-dimensional mathematical models for milling stability are estabilished.The milling stability boundary is solved to obtain the stability lobe diagram using zero-order frequency domain analysis method(ZOA method)and the semi-discretization method(SD method).The correctness of lobe diagram is verified by analyzing these two lobe diagram and comparing with the experimental data.The deformation prediction finite element simulation processing on the thin-walled parts high speed milling is analysed,taking an example of typical thin-walled parts-aluminum mirror holder bore machining.Using ABAQUS finite element analysis software,the finite element modeling technology,clamping deformation simulation,finite element calculation of the amount of deformation and other issues on the thin-walled parts are studied.On the basis of understanding of the law on the deformation,the thin-walled parts processing level cycle error compensation scheme is put forward,finally the deformation law and error compensation scheme are validated through processing test.Milling deformation is less than 0.025mm,proving that the proposed error compensation method is correct and effective.The machined surface formation mechanism on thin-walled parts high speed milling is studied and the relationship between the surface roughness Ra and the tool radius,the radial depth of cut and other processing parameters are derived.The surface roughness is predicted using adaptive neural network fuzzy inference system(ANFIS)method.The method's model structure and working principle,the establishment of the predicted models and training models are elaborated separately.The predicted models are validated through comparing with the processing test and the results by ANFIS method.The average error is within 6%,proving the prediction model is correct.A multi-objective parameters have been optimized aimed at thin-walled parts high speed milling based on genetic algorithm.Productivity,production cost and the surface roughness linearly weighted,a different weighting coefficient is set for the roughing and finishing stages.For rough machining,productivity and production costs are the main optimization goals;for finish machining,the surface roughness is a major optimization goal.Cross-encoding variables are coded according to the schema theorem.The theoretical method and testing basis are provided for choosing the optimized cutting parameters.In order to meet the high cutting rate requirement,the fuzzy control,on-line adjusting the cutting feed rate method is put forward to realize constant force in CNC high speed milling.The input variables of fuzzy controller are the difference between the reference and the actual cutting forces and the change rate of the deviation,after fuzzifying,fuzzy inference and defuzzifying,the output variable is the change of per tooth feed.The control strategy of on-line adjusting fuzzy rules and on-line self-adaptiving the output scaling factor is used in fuzzy controller.The CNC high speed milling process model is established by 3-3-5-1 type BP neural network,and the input variables are cutting parameters and the output variable is the actual cutting forces.Combining the fuzzy controller and the BP neural network,high speed milling constant force control model is constructed.By simulation and practical verification,metal removal rate is significantly increased by 18%and 29%separately.For two typical thin-walled parts-cover enclosure and the integtraed impeller,several aspects,namely,the choice of the blank and cutting tool,the design of fixture,the technological processing,the preparation and simulation of CNC programs,specimen processing and measurement,are studied,in order to achieve high efficiency,high quality,low cost thin-walled parts high speed milling.